Better analytics make for better marketing decisions. This is true whether you're talking about Facebook or billboards on the side of the road. If you don't track to see if something works, it didn't. Or maybe it did. But just like the number of licks it takes to get to the Tootsie Roll center of a Tootsie Pop, the world may never know.
Billboards are still hard to track (we will talk about them later), but digital doesn't need to mean tossing spaghetti at the wall and hoping something sticks. In an article about finding ROI in social, I made reference to campaign tagging in Google Analytics and got a lot of questions on how to use it. So, this is my guide to Google Analytics campaign tagging.
Web Analytics Basics
When a visitor goes from one website to another, a lot of data is also sent using an HTTP header called the referrer. Info like referral source (where the visitor came from), IP address, device type and screen resolution are sent in this HTTP header. Analytics packages use cookies to capture all this data. Packages like Google Analytics layer other information on top of the information sent in the header -- turning data like a visitor's IP address into a GPS location, or a search query string into keywords.
Google Analytics automatically identifies Source from the data sent, along the HTTP header. The exception is when a link from an HTTPS site leads to an HTTP site, because the encryption will not allow that information to be shared with the unencrypted website. Because of this, any referral links from an HTTPS to an HTTP site will show up as (Direct/Non) in Google Analytics. I later explain the encryption referral problem and why many HTTPS social referrals are not affected. Read that post here.
Shameless plug for my other post aside, this referral traffic problem is huge for email marketing. Think about it. Most email is, by this point, encrypted. This means your email campaigns can now be as blind as television commercials -- commercials whose tracking is determined by someone who thinks no data is better than some data. Just like billboards, we can track TV. But, we'll get to that later.
Dramatic Entrance for Campaign Tagging!
Campaign Tagging allows you to overwrite some of the default data collected by Google Analytics with your own information. This data will also be automatically added to the Acquisition>Campaign section of Google Analytics for campaign reports. Google supports the following tags: Source, Medium, Term, Content & Campaign. Theses tags are written out along with a value you assign (i.e. utm_source=newsletter). When used, these URL parameters overwrite the default data Google Analytics collects with your specified information. The example I just showed would change my referral source to newsletter.
Before I get ahead of myself, let's look at this helpful chart from Google.
You should only use tagging when it's needed and you should use only the parameters you need. You don't have to tag internal links or guest posts where you already have referral data. I know with 5 modifiers it's tempting to build very complex URLs, but most of the time that will just clutter your data. In a worst case scenario, it will make your data unusable.
In most cases, you only need Medium, Source and Campaign Name. I would view these as mandatory, although Google Analytics may work if you use only one of these parameters. Most of my tests resulted in the traffic using only utm_source or utm_medium as being placed in the Direct bucket. This is obviously not what you want out of Campaign Tagging. Regardless of what Google accepts, those three parameters will help build your automatic reports and improve your ability to analyze data by giving you nice wide segments.
Let's Talk About Each Tag
It's easy to confuse many of these tags. Be careful. If you do confuse them, you will mess up your report. Google's chart is helpful, but make sure you really understand how to use these parameters effectively.
Name
The name is pretty simple. It's used to set up your automatic reports. I use a lot of them for clients, in order to track things like seasonal sales and promotions. Since I don't have permission to use any of my clients analytics, this is a much less impactful screenshot of my site's analytics.
Some campaigns like Buffer are automatically created. Buffer ads UTM strings with the campaign name "buffer." When you, or someone else, share a link in buffer, this will show up. I'm not a fan of the way Buffer does this, but it does show their value across the internet.
Medium
For the best ability to track and perform, the analysis medium should be as broad as is logical. I like using mediums such as email, social, TV, etc... Not only do these give nice big segments to examine later, but they are not so broad as to provide almost no meaningful data as a tag such as utm_medium=internet did. Yes, I saw that in a client's analytics. In my mind, medium is the most useful part of my data because I can break down what is working by channel and, at a glance, compare email to social. Remember, medium is not the campaign name.
Source
Source is the specific referral. Think about medium and source as a soft candy within a hard candy. Social is a medium. Facebook is a source. Do you see what I mean about segments?
Term
Mostly used for PPC campaigns, term is the keyword you're bidding on. But it can be much more. You could use term to keep track of what month a newsletter is sent out.
Example: /?utm_source=newsletter&utm_medium=email&utm_term=january15&utm_campaign=engagement.
This would tell me that the medium was email and that the type of email was my newsletter. It would also tell me that the newsletter was sent in January 0f 2015, and would automatically be added to my engagement campaign for reporting.
Content
The Content parameter is to differentiate ads. I use this all the time to AB test Twitter and Facebook. I'm sure that, just like you can use term to keep track of the data for email campaigns, there are other creative uses for utm_content. For example: You could track links from different sections of your emails or different placements on a website -- even different sized ads.
Since I have yet to point it out, and I just talked about PPC, we should probably backtrack a bit. If you link your Google Analytics and Google AdWords accounts, you have an option to enable auto-tagging. Saving a lot of time manually tagging AdWords URLs because Google Analytics automatically tracks everything.
Now Let's Track a Billboard (or any offline campaign)
Let's say I have a billboard on the side of the freeway asking people to join my email list. Why I would do this? I don't know. But, obviously, I just put the URL with the UTM tracking I want (say https://www.masonpelt.com/gtml?utm_source=billboard-1&utm_medium=billboard&utm_campaign=subscribe) on the billboard. I'm going to assume that no one will remember this long, nasty URL and, therefore it will lead to no tracking data.
If I put that long URL out in the real world, be it billboard, magazine, TV, or radio, it's going to be meaningless since no one will use it. However, I could use another short URL that is only on the billboard, and have a server-side redirect from that URL to my long tagged up URL. This method is great for, magazines, newspapers and billboards. I would also recommend it for radio and TV. Redirect URL's were how I tracked my last round of radio campaigns. But you can do more.
But Wait! There's More With This Special TV Analytics Offer
Radio and TV have specific air times. (After the fact, you know when they aired. Don't rely on your media plan.) They also have defined broadcast areas. And, it's a safe assumption that most conversions will happen within 3 minutes of the air time.
This air time and viewing radius data can be added to Google Analytics with a custom data import. This lets you add a CSV of the exact air time and broadcast radius for each commercial. So, now, even without a redirect URL, you can track conversions based on direct traffic from within the viewing radiuses of the TV campaign in the time following the commercial airing.
Without the redirect URL, it won't be perfect, but it's better than no data at all. (No data would leave you with a black hole of organic traffic.) This method will at least allow you to draw a correlation between "organic referral" traffic on the website in the broadcast radius (in the three minutes following the start of the commercial) as well as the Google Analytics visits from that location and time. Unless you ask someone with absolutely no business working in marketing, "Get as much data is you can," will be a pretty normal response.
So, How Many Licks?
Some people, like the group of Purdue University engineering students, went to the effort of building a licking machine modeled after a human tongue. The Purdue students found that it took an average of 364 licks to get to the Tootsie Roll center of a Tootsie Pop.
This number was disputed by a University of Michigan chemical engineering PHD student. University of Michigan made their own licking machine and determined that it required 411 licks to reach the chewy, chocolate, candy center.
Back at Purdue, 20 students took on the the licking challenge, without the aid of a cyborg's tongue. They averaged 252 licks each to reach the center. (Source)
These are the people who you want on your analytics team! Analytics is a science. It's much more detailed than handing a product to someone who professes to be wise and having the crazy hoot owl make a claim as wild as saying that it only takes three licks to make it through the hard candy shell. It's about designing experiments to test what works. Even when it's inconclusive, you can take the information, apply it, and build on it. Again, some data is better than no data.
If you have a better way to track offline to digital conversions, or any cool thoughts and suggestions for using campaign tagging, let me know in the comments.
(image: What is inside a tootsie pop // by: theilr)